Metallicities of the Milky Way globular cluster (GC) system are interpreted within the framework of a closed box model of chemical evolution. Model age-metallicity relations normalised to the Sun are compared with an observed age-metallicity relation constructed from literature data for the Galactic globular clusters and solar neighbourhood dwarf stars. One branch of the GC age-metallicity relation is seen to form a natural extension of the solar neighbourhood relation to low metallicities. Such clusters may fit in situ into the same Galactic closed-box model as many disk stars. Timescales for halo and disk GC formation are computed based on this assumption. However, a single closed box cannot account for a second branch in the GC age-metallicity data. This problem is addressed by assuming that some GCs did not form in situ within the closed box of the Galaxy, but rather were formed within separate parent objects whose chemical evolution was characterised by longer star formation timescales than the Milky Way box. These parent objects were eventually acquired by the Milky Way. Estimates are made of the initial mass and star formation timescale of those objects that contributed GCs to the accreted component of the Milky Way.

银河系球状星团（GC）系统的金属在化学演化的密闭盒模型框架内解释。将规范化到太阳的模型年龄-金属性关系与从银河系球状星团和太阳邻区矮星的文献数据构建的观测到的年龄-金属性关系进行了比较。可以看出，GC年龄-金属性关系的一个分支自然形成了太阳邻域关系向低金属性的自然延伸。这样的星团可能就地适合与许多磁盘星一样的银河密闭盒模型。根据此假设计算晕圈和磁盘GC形成的时间标度。但是，一个封闭的框不能解释GC时效金属数据中的第二个分支。通过假设某些GC不是在银河系的密闭盒中原位形成，而是在单独的父对象中形成的，解决了这些问题，这些对象的化学演化特征是比银河系更长的恒星形成时间尺度。这些父对象最终被银河系捕获。估计那些将GC贡献给银河系增加部分的天体的初始质量和恒星形成的时间尺度。